up and down the slope, at an average of about four
feet depth, going five feet deep on the brow of the hill, to cut
through the brush-drain. I used two-inch sole-tiles for minors, and
three-inch for the main.

The effect was instantaneous. The land which, in the Spring of
1856, had been so wet that it could not, even though partially
drained with brush-drains, be planted till the 5th of June, was, in
1857, ready to work as soon as the snow was off. My farm journal
says, under date of April 6th, "plowed drained land with double
plow two days after a heavy storm--dry enough." I spent that Summer
in Europe. The land was planted with corn, which produced a heavy
crop. I find an entry in my journal, on my return, "My drained land
has been in good condition--neither too wet nor too dry--all
Summer."

In the Fall of 1857, I laid about 170 rods in other parts of the
field, at similar depths and distances, and in 1858 completed the
upper part, on which is an orchard of apple trees. A part of this
orchard was originally so wet as to kill the trees the first year,
but by brush-drains I dried it enough to keep the next set alive.
There was no water visible at the surface, and the land was dry
enough for corn and potatoes; still the trees looked badly, and
many were winter-killed. I had learned the formation of the earth
about my premises, of which I had at first no adequate conception,
and was satisfied that no fruit tree could flourish with its feet
in cold water, even in Winter. All nursery-men and fruit-growers
agree, that land must be well drained for fruit. I therefore laid
four-foot tile drains between the rows of trees, in this apparently
dry sand. We found abundance of water, in the driest season, at
four feet, and it has never ceased to flow copiously.

I measured accurately the discharge of water from the main which
receives the drainage of about one and a half acres of the orchard,
at a time when it gave, what seemed to me an average quantity for
the Winter months, when the earth was frozen solid, and found it to
be about 480 barrels per day! The estimate was made by holding a
bucket, which contained ten quarts, under the outlet, when it was
found that it would fill in fifteen seconds, equal to ten gallons
per minute; and six hundred gallons,